I'll be upfront: I'm building AI-powered tools for change management, so I have a stake in this conversation. But I also have 20 years of practitioner experience — which means I'm not going to tell you AI replaces the human work. It doesn't.
What it does is change which parts of the job can be done faster, smarter, and at scale. And the stakes are high. Between 70 and 85% of AI projects fail to deliver their expected benefits — a failure rate twice that of traditional IT projects, according to BCG. The primary causes aren't technical. They're organisational: poor data, unclear objectives, and — you guessed it — insufficient change management. Meanwhile, 42% of companies abandoned most of their AI initiatives in 2025, up from 17% the year before. Organisations are spending billions on AI. They are not spending nearly enough on the people side of it.
- Diagnosis at scale
- Communication scaffolding
- Research synthesis
- Reading the room
- Building trust
- Navigating conflict
- Knowing when to push — and when to hold
What AI is genuinely good at in change
1. Diagnosis at scale
Historically, a change impact assessment meant weeks of workshops, interviews, and a consultant synthesising outputs into a report that was often out of date by the time it landed. AI compresses that cycle dramatically — pulling patterns from structured inputs and generating team-level insights in minutes. Not perfect, but directionally accurate and far faster than the alternative.
2. Communication scaffolding
Change managers spend an enormous amount of time drafting communications that follow the same structural logic over and over. AI handles that scaffolding. The practitioner adds political awareness, organisational voice, and the judgement that only someone inside the room can bring.
3. Research synthesis
The change management evidence base is vast. AI can surface relevant frameworks, research, and risk factors based on the specific context of a project — so practitioners spend less time searching and more time applying.
What AI is not good at
Reading the room. No model can tell you that the CFO's silence in the last steering committee means she's about to pull funding. It can't hear the tone shift in a team meeting, or notice that a particular leader has gone quiet. A 2025 survey found that 89% of workers are concerned about AI's impact on their job security. That anxiety is invisible to a model unless a human surfaces it. The relational work of change — the trust-building, the conflict navigation, the knowing when to push and when to hold — that remains irreducibly human.
What this means for the profession
The practitioners who will thrive are the ones who adopt AI as a force multiplier for the diagnostic and documentation work, and redirect their human time to the things AI cannot do. Kyndryl research found that 45% of CEOs say their employees are reluctant or hostile toward AI adoption. That's not a technology problem. That's a change problem — and it needs a change manager, not a prompt engineer.
AI is not the change manager of the future. It's the tool that frees the change manager to do the job properly.
Find out how ready your teams are
If you're rolling out AI and you want to know how ready your teams actually are, start with the Change Made Simple Scorecard. It's free, it takes minutes, and it gives you a team-by-team risk picture before things go sideways.
